Why AI investing is nowhere near its peak
Alex Pollak built Loftus Peak into a $1.5 billion fund on disruption investing – now he’s telling Henry Jennings exactly where AI is taking it next.
Henry Jennings recently sat down with Alex Pollak, founder and CIO of Loftus Peak, to talk AI, disruption investing, and where the smart money is moving next. Loftus Peak manages $1.5 billion and has been running its disruption thesis for 12 years – long before AI became a household word.
Why disruption still has legs
Loftus Peak doesn’t call itself a technology fund. It calls itself a disruption fund, and the distinction matters. The thesis traces back to the way Seek, carsales.com, and realestate.com.au dismantled Fairfax’s classified advertising business – taking something that lived on the back of a newspaper and making it native to the internet. From there, the same logic applied to retail, cloud computing, social media, and ultimately AI.
Every industry is subject to disruption given sufficient bandwidth, and the fund has simply followed that thread from one sector to the next. Nvidia (NASDAQ: NVDA) was on Loftus Peak’s radar in 2016 – well before the AI trade became consensus – identified as one of the first genuine machine learning companies.
The AI bubble question
Is AI a bubble? Pollak’s answer is nuanced. The dot-com crash happened because burn rates outpaced any credible revenue model. The internet was narrowband, the economics didn’t work, and companies like Pets.com were spending money that demand could never justify. This time the demand is real and the monetisation is demonstrable – Anthropic and OpenAI are running at tens of billions in annual revenue. The product works.
The open question – and Pollak is candid that nobody actually knows the answer – is whether the current level of capital investment is proportionate to that demand. A trillion dollars deployed in a year is roughly equivalent to the entire US annual interest bill on its $39 trillion of debt. That is not a small number. But with Alphabet (NASDAQ: GOOGL), Microsoft (NASDAQ: MSFT), and Amazon (NASDAQ: AMZN) still in net cash positions despite the spending, and Nvidia on track to return half a trillion dollars to shareholders from free cash flow over the next two or three years, the underlying numbers are harder to dismiss than they were in 1999.
Where Loftus Peak is putting money to work
Beyond Nvidia – the fund’s largest position – Pollak identified three areas where Loftus Peak is actively building exposure.
China is the biggest move. The fund holds positions in Tencent (HKEX: 700), Alibaba (NYSE: BABA), CATL (SZSE: 300750), and BYD (HKEX: 1211), with China weighting already above 10% and likely to climb significantly. These aren’t pure AI plays – they’re companies that use AI as a tool – but Pollak sees Chinese technology and battery stocks as genuinely underappreciated by most Australian investors.
Biotech is the second area, anchored by Eli Lilly (NYSE: LLY) and its weight loss drug franchise. Pollak’s view is that solving obesity isn’t just a pharmaceutical milestone – it removes the preconditions for a cascade of other conditions, including certain cancers, liver disease, heart disease, and kidney disease. The fund is also tracking developments in solid tumour oncology, where remission rates for melanomas are now reaching three to five years in clinical studies – outcomes that were unheard of five years ago.
Streaming rounds out the three, with positions in Spotify (NYSE: SPOT), Netflix (NASDAQ: NFLX), and Roku (NASDAQ: ROKU). Pollak describes it as a chunky bet that is getting larger.
The picks and shovels approach
Across all of these positions, Loftus Peak’s AI exposure is deliberately structured around infrastructure rather than applications. The fund isn’t trying to pick which transport company uses AI best – it’s owning the companies that build the tools those transport companies buy. The goal is broad-based validation of AI across industries, not concentration in a handful of platform names.
Best ideas right now
When pushed for his single best idea, Pollak went to two. The first is CATL. The electric vehicle battery business is large, but Pollak’s conviction is that the energy storage and solutions market – batteries for homes, data centres, ships, and eventually aircraft – is at least as large and probably multiples of it. CATL is not expensively valued relative to that opportunity.
The second is Eli Lilly. Pollak’s framing is straightforward: a pill that solves obesity and the cascade of conditions that follow from it represents one of the most significant reductions in human disease burden in modern medicine. That is a long runway.
On SpaceX and what retail investors should do
Loftus Peak has not participated in SpaceX’s private rounds and has no current plans to. Pollak’s view is that on current metrics the valuation doesn’t make sense as an investment – though he left the door open depending on IPO pricing when a live transaction exists.
For retail investors looking at the tech sector and wondering whether they’ve missed it, Pollak’s honest answer is that you can still make money – but this is a market that rewards concentrated research. The fund runs eight or nine analysts, travels to the US, Taiwan, and China, and uses serious financial modelling tools to understand what’s going on inside these companies. That concentration of research is the edge. Without it, the better answer is to let someone who does that work full-time do it for you.